L’ADAPTATION À LA DEMANDE PSYCHOLOGIQUE AU TRAVAIL : LE RÔLE DE L’AUTODÉTERMINATION ET DE L’AUTONOMIE DÉCISIONNELLE AFIN
DE PREDIRE L’EPUISEMENT PROFESSIONNEL
Mémoire présenté
à
la Faculté des études supérieures de l’Université Laval
pour l’obtention du grade de maître en psychologie (M.Ps.)
École de psychologie
FACULTÉ DES SCIENCES SOCIALES UNIVERSITÉ LAVAL
JUIN 2002
La présente étude a pour objectif d’étudier les différences individuelles inhérentes à !’utilisation de l’autonomie décisionnelle dans un travail où la demande psychologique est élevée et ce, afin de prédire l’épuisement professionnel. Un triple effet d’interaction est alors démontré entre la demande psychologique, l’autonomie décisionnelle et !'autodétermination au travail comme prédicteurs des dimensions de l’épuisement professionnel (l’épuisement émotionnel, la dépersonnalisation et 1 ’accomplissement personnel). De fait, un niveau élevé d'autonomie décisionnelle atténue la relation entre la demande psychologique et les dimensions négatives de l'épuisement professionnel (l’épuisement émotionnel et la dépersonnalisation) chez les employés fortement autodéterminés. De plus, un niveau élevé d'autonomie décisionnelle augmente la relation entre la demande psychologique et le sentiment d'accomplissement personnel uniquement chez les employés fortement autodéterminés. En conséquence, les résultats apportent certaines nuances au modèle Demande-autonomie au travail, en démontrant le rôle modérateur de 1 ’autodétermination dans la prédiction de l’épuisement professionnel.
Frédéric Guay, Ph D. êngg%Th.D.
Ce mémoire ainsi que l’aventure universitaire s’y rattachant représente pour moi 1 ’accomplissement d’un objectif personnel important. Au cours de ce périple, certaines personnes, chacune à leur façon, ont contribué à la réalisation de ce projet. Je profite donc de cette occasion pour les remercier du fond du cœur.
Tout d’abord, je désire exprimer ma reconnaissance envers mes directeurs de mémoire Mme Caroline Senécal et Monsieur Frédéric Guay. Leurs conseils judicieux, leur grande disponibilité, leur soutien constant ainsi que leur passion pour leur travail en font des directeurs remarquables. Je tiens également à les remercier pour la confiance qu’ils me témoignent, notamment en me permettant de collaborer à leurs projets de recherche.
Je désire aussi remercier mes collègues de travail du laboratoire de psychologie sociale, Étienne, Lysanne et Marie-Noëlle, qui ont contribué par leur présence et leur partage à créer un climat de travail fort agréable. Je ne peux passer sous silence le
soutien de mes coéquipiers de hockey qui m’offrent l’opportunité de sortir de mon travail intellectuel en me permettant de pratiquer un sport que j’adore.
Je tiens finalement à exprimer ma gratitude envers ma famille. À ma mère
Claudette et mon père René, pour leur amour et leur présence ainsi qu’à ma sœur Mylène et mon frère André, pour leur encouragement et l’intérêt qu’ils portent à la réussite du bébé de la famille. Merci à ma compagne, Stéphanie, non seulement pour sa présence et ses multiples encouragements, mais également pour sa grande générosité et son appui qui font d’elle un être tout à fait spécial.
ii iii iv vi vii 1 RÉSUMÉ AVANT-PROPOS
TABLE DES MATIÈRES LISTES DES TABLEAUX LISTES DES FIGURES
INTRODUCTION GÉNÉRALE 4 5 6 7 7 8 14 17 17 17 21 21 ARTICLE : “Adjusting to Job Demands: The Role of Work Self-Determination and Job Control in Predicting Burnout”
RÉSUMÉ ABSTRACT INTRODUCTION
Burnout
The Job Demand-Control model The role of individual characteristics METHOD
Participants and procedures Measures
RESULTS
Preliminary analyses
21 Regression analyses
DISCUSSION 24
Theoretical implications 26
Limitations and future studies 28
Conclusion 29 Note 30 TABLES 31 FIGURES 35 CONCLUSION GÉNÉRALE 40 RÉFÉRENCES 43
Tableaux
1 Correlations among all variables and descriptive statistics 31 2 Hierarchical multiple regression analyses for emotional exhaustion 32 3 Hierarchical multiple regression analyses for depersonalization 33 34 Hierarchical multiple regression analyses for personal accomplishment
36
37
38
39 Figures
1 The Job Demand-Control model
2 Three-way interaction between job demands, job control and self- determined work motivation when predicting emotional exhaustion 3 Three-way interaction between job demands, job control and self-
determined work motivation when predicting depersonalization 4 Three-way interaction between job demands, job control and self-
dernières se manifestent par une complexification des tâches, un changement rapide et une compétitivité accrue (Goldstein, 1993; Smith, Ford, & Kozlowski, 1997; Cascio, 1998). Ces changements, dans F organisation du travail, provoquent une surcharge de travail qui peut créer des problèmes de santé psychologique et physique chez plusieurs employés (DeFranck & Ivancevich, 1998).
Au Québec, certaines études ont démontré que des niveaux élevés de stress au travail deviennent plus prévalants et qu’ils peuvent engendrer des conséquences négatives pour la santé psychologique et physique (Vézina, 1999). Parmi les problèmes de santé psychologique les plus couramment cités, on retrouve l’épuisement professionnel. De fait, le travail est devenu moins exigeant au plan physique, mais davantage contraignant au plan psychologique et émotionnel (Marmot, Siegrist, Theorell, & Feeney, 1999). Plusieurs modèles théoriques ont émergé afin d’expliquer les liens entre !’organisation du travail et l’épuisement professionnel. Parmi ceux-ci, on retrouve le modèle Demande-autonomie développé par Karasek (1979; Karasek &
Theorell, 1990). Ce modèle propose que l’autonomie décisionnelle protège l’individu des effets néfastes associés à !’organisation du travail. Bien que ce modèle s’avère intéressant pour comprendre le lien qui existe entre !’organisation du travail et l’épuisement professionnel, il ne tient pas compte des différences individuelles pour expliquer un tel lien. Pourtant, de nombreuses études en psychologie organisationnelle ont démontré que les gens ne réagissent pas tous de la même manière vis-à-vis de !’organisation du travail. En effet, certains individus paraissent bien s’adapter aux situations stressantes alors que d’autres éprouvent des difficultés (Parkes, 1990,
L’objectif de ce mémoire consiste à vérifier si certaines caractéristiques individuelles (Le., 1 ’autodétermination au travail) peuvent expliquer les liens entre !’organisation du travail et
l’épuisement professionnel. L’étude met l’accent sur deux dimensions cruciales de !’organisation du travail, soit la demande psychologique et l’autonomie décisionnelle. D’une part, la demande psychologique réfère à la quantité de travail à accomplir, de même qu’aux exigences et aux
contraintes de temps liées au travail. Quant à l’autonomie décisionnelle, elle représente la maîtrise du processus du travail. Elle réfère à la capacité de prendre des décisions et à la possibilité
d’exercer un certain contrôle sur le travail à réaliser. Ainsi, selon le modèle Demande-autonomie au travail (Karasek, 1979; Karasek & Theorell, 1990), la demande psychologique au travail peut mener à deux conséquences opposées, soit la tension au travail ou le développement des
apprentissages au travail. Selon ce modèle, le type de conséquences vécues par l’employé varie en fonction du niveau d’autonomie décisionnelle que présente !’organisation du travail. En effet, une organisation de travail exigeante et offrant peu d’autonomie décisionnelle à l’employé accentue le risque que celui-ci développe de la tension au travail, pouvant mener ainsi à des problèmes
d’épuisement professionnel. À l’inverse, une organisation de travail exigeante qui soutient l’autonomie décisionnelle favorise le sentiment de maîtrise de l’employé et lui offre l’opportunité de résoudre des problèmes. En conséquence, un employé bénéficiant d’une telle organisation du travail vit moins de tension au travail et par surcroît, développe davantage ses habiletés vis-à-vis du travail à accomplir.
La présente étude se propose donc d’étudier les différences individuelles inhérentes à !’utilisation de l’autonomie décisionnelle dans un travail où la demande psychologique est élevée et ce, afin de prédire l’épuisement professionnel. À cette fin, nous vérifierons les interactions qui existent entre la demande psychologique, l’autonomie décisionnelle et la personnalité de
l’employé, soit son degré d’autodétermination au travail, afin de prédire les trois principales dimensions de l’épuisement professionnel (Le., l’épuisement émotionnel, la dépersonnalisation et le sentiment d’accomplissement).
Cette étude s’avère donc novatrice pour trois raisons. Premièrement, elle évalue le rôle d’une caractéristique personnelle, soit !’autodétermination au travail, variable qui n’a pas encore été étudiée pour comprendre les relations qui existent entre l’autonomie décisionnelle, la demande psychologique et l’épuisement professionnel. Deuxièmement, au meilleur de nos connaissances, aucune étude n’a démontré le rôle de la personnalité (Le., 1 ’autodétermination au travail) dans la prédiction de chacune des trois dimensions de l’épuisement professionnel (Le., l’épuisement émotionnel, la dépersonnalisation et le sentiment d’accomplissement). Troisièmement, très peu d’études ont mis à l’épreuve l’hypothèse du développement des apprentissages au travail du modèle de Karasek et ce, même si ce type d’apprentissage peut retarder ou même enrayer le développement des problèmes de santé au travail (Theorell & Karasek, 1996). De plus, les études qui se sont attardés au développement des apprentissages n’ont pas évalué le rôle des différences individuelles.
Le présent mémoire est rédigé sous forme d’article scientifique et a été soumis au Journal of Vocational Behavior. La première section de l’article fait état de la recherche sur l’épuisement professionnel et le modèle Demande-autonomie au travail. Ensuite, elle présente le rôle des caractéristiques individuelles, les objectifs spécifiques et les hypothèses de l’étude. La deuxième section expose la méthodologie utilisée. La troisième section énonce les résultats obtenus lors de l’étude. Enfin, la dernière section présente la discussion portant sur les implications théoriques des résultats obtenus ainsi que les limites de l’étude et propose des suggestions de recherches futures.
Adjusting to Job Demands:
The Role of Work Self-Determination and Job Control in Predicting Burnout
Claude Fernet, Caroline Senécal, Frédéric Guay Université Laval
Date : May 22, 2002
Author’s notes
Claude Fernet, École de Psychologie; Caroline Senécal, École de Psychologie; Frédéric Guay. Département des fondements et pratiques en éducation. We would like to thank Caroline Biron and Jean-Pierre Brun for their contribution to this research project. Correspondence concerning this article should be addressed to Claude Fernet, École de Psychologie Université Lavai, Ste-Foy, Québec, G1K 7P4. Phone: (418) 656-2131 ext. 8437, Fax: (418) 656-3646. Electronic mail may be sent to: Labocs@psy.ulaval.ca.
Résumé
La présente étude a pour objectif d’étudier les différences individuelles inhérentes à !’utilisation de l’autonomie décisionnelle dans un travail où la demande psychologique est élevée et ce, afin de prédire l’épuisement professionnel. Un triple effet d’interaction est alors démontré entre la demande psychologique, l’autonomie décisionnelle et !'autodétermination au travail comme prédicteurs des dimensions de l’épuisement professionnel (l’épuisement émotionnel, la dépersonnalisation et 1 ’accomplissement personnel). De fait, un niveau élevé d'autonomie décisionnelle atténue la relation entre la demande psychologique et les dimensions négatives de l'épuisement professionnel (l’épuisement émotionnel et la dépersonnalisation) chez les employés fortement autodéterminés. **De plus, un niveau élevé d'autonomie décisionnelle augmente la relation entre la demande psychologique et le sentiment d'accomplissement personnel uniquement chez les employés fortement autodéterminés. En conséquence, les résultats apportent certaines nuances au modèle Demande-autonomie au travail, en démontrant le rôle modérateur de l’autodétermination dans la prédiction de l’épuisement professionnel.
a high job demands context in order to predict burnout. A three-way interactive effect has been demonstrated between job demands, job control and self-determined work motivation predicting each dimension of burnout (emotional exhaustion, depersonalization and personal accomplishment). A high level of job control attenuates the relation between job demands and the negative dimensions of burnout (emotional exhaustion and depersonalization) among high self- determined employees. A high level of job control increases the relation between job demands and the sense of personal accomplishment only in high self-determined employees. The results add certain nuances to the Job Demand-Control model by demonstrating the moderating role of self-determination in predicting burnout.
Key words: work self-determination; job demands; job control; burnout; individual characteristics; work adjustment; motivation; occupational stress and strain.
Adjusting to Job Demands:
The Role of Work Self-Determination and Job Control in Predicting Burnout
In recent years, many researchers have attempted to explain, using theoretical models, the psychological well-being of workers in relation to work environment (see Cooper, 1998; Parker & Wall, 1998). Among these theoretical models, the Job Demand-Control model (Karasek,1979, 1998; Karasek & Theorell, 1990) has been widely studied. This model suggests that job control protects the individual from the unhealthy effects related to work environment. Although this model proves to be useful for understanding the link that exists between work environment and work adjustment, the latter does not take into account individual differences to explain such a link. However, numerous studies in organizational psychology have demonstrated that not all people react in the same way to work environment (Parkes, 1990, 1994). In fact, it would seem that some individuals adjust well to stressful situations while others have problems adjusting. The aim of this study is to determine whether certain individual characteristics (i.e., self-determined work motivation) can explain the links that exist between job demands, job control and burnout (i.e., emotional exhaustion, depersonalization and personal accomplishment). In the sections that
follow, we will examine the following elements: a) burnout, b) the Job Demand-Control model, c) the role of individual characteristics and d) the specific goals and the hypotheses of this study. Burnout
Most authors agree on the three-dimensional construct of burnout (Kop, Euwena, & Schaufel!, 1999; Schaufel!, Maslach, & Marek, 1993). Burnout is defined as a symptom of emotional exhaustion, depersonalization and reduced personal accomplishment at work due to work activity (Maslach, 1982). Emotional exhaustion refers to the depletion of one’s emotional resources. Depersonalization refers to the shell into which employees withdraw to protect
themselves from psychological stress coming from people with whom they interact at work. This shell is expressed as a detached attitude toward the human characteristics of other employees at work. Reduced personal accomplishment refers to a decrease in the feeling of competence and productivity at work. Thus, in addition to being marked by a loss of emotional energy, burnout implies a negative assessment of self (reduced personal accomplishment) and of others
(depersonalization). Studies suggest that the three dimensions of burnout are associated with different aspects of work environment, in particular work overload and role ambiguity or conflict (Janssen, Schaufeli, & Houkes, 1999). Moreover, some studies underline the role of individual characteristics in individuals’ vulnerability to developing burnout (Semmer, 1996). However, currently, studies on burnout tend to consider individual and contextual factors separately rather than using a theoretical framework which incorporates them together (i.e., job-person fit). According to Maslach, Schaufeli, and Leiter (2001), the challenge is to extend the job-person paradigm to a broader and more complex conceptualization of the person situated in the job context.
Job Demand-Control model
One of the models guiding the research on burnout is the Job Demand-Control model developed by Karasek (1979, Karasek & Theorell, 1990). This model focuses on two dimensions found in work environment, i.e., job demands and job control. Job demands refer to the volume of work to be accomplished, as well as the requirements and time constraints related to the work. Job control refers to the control over work process, that is, the ability to make decisions and the opportunity to exercise a degree of control over the work to be accomplished.
This model suggests that job demands can have two opposing consequences: 1) job strain or 2) learning-oriented outcomes at work (see Figure 1). Thus, job demands do not have the same
impact on all employees. For example, as a result of high job demands, some employees feel distressed while other employees feel challenged. Based on this model, the type of consequences experienced by the employee varies according to the level of job control involved in the work environment. In fact, a work environment that is demanding and offers little job control to the employee would increase the employee’s risk of developing job strain that can lead to physical and psychological health problems. Conversely, a demanding job (but not overwhelming) that offers job control would increase the employee’s feeling of control and opportunity to solve problems. Therefore, incumbents of this type of job (i.e., active job) would experience less job strain and, moreover, increase learning and development of theirs skills.
Thus, the model is based on the premise that job strain and active learning at work are derived from the interactive effect between job demands and job control (Ganster, 1989). However, very few of the studies that have explicitly tested the interaction between these two components have supported the Job Demand-Control model (see de Jonge & Kompier, 1997, Van der Doef & Maes, 1999, for a recent review of the literature). Moreover, a number of studies have demonstrated that not all employees respond favourably to job control (e.g., de Rijk, LeBlanc, Schaufeli, & de Jonge, 1998; Parker & Sprigg, 1999; S alano va, Peiro, & Schaufel!, 2002;
Schaubroeck & Merritt, 1997; Schaubroeck, Jones, & Xie, 2001). Among certain employees, job control can actually exacerbate the psychological strain experienced at work. Furthermore, very few studies have successfully demonstrated that the interaction between the employee’s job demands and job control favours learning-oriented outcomes (de Jonge & Kompier, 1997; Karasek, 1998).
The present study aims to examine the individual differences inherent to the use of job control in a high job demands context in order to predict burnout. Thus, we will verify the
interactions that exist between job demands, job control and employees’ personality, ( i.e., their degree of self-determined work motivation), in order to predict each dimension of burnout (i.e., emotional exhaustion, depersonalization and personal accomplishment). This study is innovative for three reasons. First, it assesses the role of a personal characteristic, ( i.e., self-determined work motivation) which has not yet been examined in order to understand the relations that exist
between job demands, job control and burnout. Second, to the best of our knowledge, no study has yet demonstrated the role of personality (i.e., self-determined work motivation) in predicting each of the three dimensions of burnout (i.e., emotional exhaustion, depersonalization and
personal accomplishment). Third, very few studies have tested Karasek’s hypothesis of learning- oriented outcomes, even though this type of learning can delay or even curb the development of occupational health problems (Theorell & Karasek, 1996). Moreover, studies which have assessed learning-oriented outcomes have not assessed the role of individual differences. Empirical evidence on the Job Demand-Control model
Initially, empirical studies tested the hypotheses of Karasek’s model (1979) by focusing on the main effects of job demands and job control to explain mental health problems. Research results showed that job control is negatively associated with various psychological health problems, including burnout (Landsbergis, 1988; Melamed, Kushnir, & Meir, 1991; Nijhuis, & Smulders, 1996; Sonnentag, Brodbeck, Heinbokel, & Stolte 1994) while the variable of job demands is positively associated with these variables. Moreover, a number of studies have demonstrated that job demands is negatively linked with learning-oriented outcomes while job control is positively linked with this variable (see de Jonge & Kompier, 1997). As regards personal accomplishment at work (cf., Maslach, 1998), a study conducted by Dollard, Winefield, Winefield and de Jonge (2000), shows that this variable is positively associated with job demands
and also with job control. However, empirical studies of this type do not allow us to ascertain whether the effects (i.e., negative or positive) of work environment can be exclusively attributed to job demands or to the degree of job control (Alfredsson, Spetz, & Theorell, 1985; Hammar,
Alfredsson, & Theorell, 1994). In this respect, several authors have suggested that it is theoretically and practically necessary to explicitly test the interaction terms (i.e., buffering hypothesis; see Beehr, Glaser, Canali, & Wallwey, 2001; Kasl, 1996; Landsbergis et al., 1994). However, studies which tested the model’s terms of interaction in relation to health or
psychological health problems (e.g., anxiety, depression, job satisfaction and burnout) are somewhat inconclusive (see Van der Doef & Maes, 1999). As regards burnout, no study has demonstrated the interactive effect between job demands and job control (see Rafferty, Friend, & Landsbergis, 2001). However, several studies on different workplaces and on different
occupations have demonstrated the main effects of job demands or job control on burnout. Karasek’s model (1979) was challenged on the theoretical level because it did not take individual differences into account. According to Parkes (1994), personality characteristics can moderate the relation between job demands and job strain in two ways: (1) a good fit between personal characteristics and work-environment characteristics results in favourable psychological consequences while (2) a lack of fit leads to negative consequences. The current state of
knowledge suggests that certain individual characteristics (e.g., self-efficacy, explanatory style, coping style and proactive personality) would influence the individual’s psychological adjustment to the constraints of work environment (de Rijk et ah, 1998; Parker & Sprigg, 1999; S alano va, et al., 2002; Schaubroeck & Merritt, 1997; Schaubroeck et al., 2001). Therefore, it can be assumed that some people would be more effective than others at managing job control. As regards burnout, two studies have examined the role of individual characteristics. First, de Rijk et al.
(1998) report that the active coping style (i.e., actively controlling the stressor) moderates the demand-control relation in the prediction of emotional exhaustion. The active coping style refers to the cognitive analysis of a situation and a concrete action undertaken to solve a problem at work. Thus, when job demands is high, employees with an active coping style experience less emotional exhaustion when they have a high level of job control. However, a high level of job control increases emotional exhaustion due to job demands in people with a passive coping style. Second, S alano va et al. (2002) demonstrate the moderating role of self-efficacy in the demand- control relation in the prediction of two dimensions of burnout (i.e., emotional exhaustion and depersonalization) among information technology workers. Results show that job control (i.e., method control and timing control) reduces the symptoms of emotional exhaustion and
depersonalization among employees who have a high level of computer self-efficacy. On the other hand, job control does not reduce the effect of job demands on these two dimensions of burnout for workers who have a lower level of computer self-efficacy.
To sum up, these studies converge on the idea that the individual’s personality is important in understanding the relations that exist between job demands, job control and certain dimensions of burnout. Job control reduces the negative effects of job demands on burnout only for people who have personal coping characteristics. On the other hand, job control does not reduce the negative effects of job demands on burnout for people who have negative personal characteristics (e.g., low self-efficacy, passive coping style). It can even contribute to the development of health problems.
Thus, a number of studies have demonstrated the importance of the role of personality in understanding the interaction between job demands and job control when predicting burnout. They therefore add to the psychological strain hypothesis of Karasek’s model (1979) by
incorporating the role of individual characteristics. However, very few studies have assessed individual differences in the prediction of learning-oriented outcomes. As regards burnout, Dollard et al. (2000) demonstrated only the main effects of job demands and job control in the prediction of personal accomplishment. Beyond the role of individual characteristics, the links with job demands and job control in relation to learning-oriented outcomes have been under- examined (Karasek, 1998; Theorell & Karasek, 1996). In fact, researchers focusing on this issue have demonstrated, at most, the main effects of job demands and job control. To the best of our knowledge, no study has demonstrated the role of personality as a moderator in the demand- control relation in predicting learning-oriented outcomes. Although Parker and Sprigg’s study (1999) demonstrated that proactive personality facilitates learning-oriented outcomes, the three- way interactive effect between job demands, job control and personality could not be
demonstrated.
All the studies reviewed bring out the importance of individual characteristics in
explaining the relations that exist between job demands, job control and burnout. However, the variables used in order to assess the individual characteristics have some limitations which may explain why the hypothesis of active learning was not supported in previous studies. Among other things, while proactive personality, explanatory style and self-efficacy underline the intentionality of behaviours, they do not allow us to differentiate between self-determined and non self-
determined behaviours, which, according to Deci & Ryan (1985), is a fundamental distinction. Self-determination, on the other hand, refers to the experience of choice in the process of intentionality of behaviours, thus going beyond the perception of a contingency between an exhibited behaviour and the understanding of any result. Several studies demonstrate that this
distinction is crucial (see Blais, Hess, Bourbonnais, Saintonge, & Riddle, 1995; O’Connor & Vallerand, 1994; Pelletier, Vallerand, Green-Demers, Brière, & Blais, 1995).
Self-Determination Theory
Self-determination theory (Deci & Ryan, 1985, 1991; Ryan & Deci, 2000) thus allows us to compensate for this limitation. More precisely, this theory states that inherent personal
resources are important for the development of a person’s personality and the regulation of behaviours. It is based on the premise that each individual has a fundamental need to experience autonomy, competence and relatedness (Deci & Ryan, 1985, 1991). Consequently, the more frequent or generalized these experiences are in the important areas of life, the more they
influence the individual’s quality of life and health. These experiences are modulated according to the dynamic interaction between the persons’ characteristics and those of their environment.
Various experiences contribute to the development of different types of motivation that correspond to distinct levels of self-regulation and integration of self, which in return, produce distinct cognitive, affective and behavioural reactions. According to Self-Determination Theory (Deci & Ryan, 1985, 1991), individuals who perform their work by choice, for pleasure and the satisfaction of performing various tasks demonstrate a self-determined behaviour regulation. On the other hand, individuals who perform their work through internal and/or external pressures demonstrate a non-self-determined regulation. In recent decades, empirical studies have confirmed that self-determined motivation was a very useful concept for understanding human behaviours in different areas of life including work (Blais et ah, 1995; Blais, Lachance, Brière, Riddle, & Vallerand, 1993; Senécal, Vallerand, & Guay, 2001).
Self-determined work motivation is a distinct concept of job control. Job control is a contextual variable that refers to the notion of control offered to the individual by his work
environment. Self-determination is an individual variable that refers to the individual’s experience of free choices (freedom of choice). Thus, there is a fundamental difference between the
possibility to choose and the will to exercise one’s choices.
A study by O’Connor and Vallerand (1994) stresses the importance of self-determination in understanding how people deal with the autonomy offered by their environment. Results of this study demonstrate that self-determined people adjust better psychologically to an environment that provides autonomy, while the adjustment of people with non-self-determined motivation is
favoured in a controlling environment that offers fewer opportunities for choice and autonomy. Moreover, they underline the need to understand the fit between personal needs and environmental characteristics. This understanding is very important while interpreting the Job Demand-Control model. In fact, as the workers’ resources and intrinsic abilities in their adjustment to job demands are crucial, they are likely to foster or delay the appearance of health problems (Caplan, 1985). The variable of job demands therefore does not have direct pathogenic effects on the workers because it is modulated by their individual characteristics. Thus, individuals would experience job strain or feel that they have attained learning-oriented outcomes, if and only if their individual characteristics (i.e., self-determined work motivation) are compatible or incompatible with work environment (i.e., job demands and job control).
Characteristics, objectives and hypotheses of the present study.
This study responds to various methodological and theoretical criticisms which could explain the inconsistent results of the Job Demand-Control model. First, it was conducted on a specific group of employees, i.e., university professors. According to Kristensen (1995), it is crucial to study the model, using a specific sample that provides a degree of variability for each of the variables examined. Second, the measures of job demands and job control used do not include
items that reflect affective elements. In fact, according to Wall et al. (1996), independent
variables which contain affective elements can lead to contrived effects on the dependent variable, when the latter represents an affective measure.
The first objective of the study is to assess the role of self-determined work motivation as a variable that moderates the demand-control relation in the prediction of two components of
burnout, i.e. emotional exhaustion and depersonalization. The first hypothesis postulates that there is a three-way interaction between job demands, job control and self-determination in predicting emotional exhaustion and depersonalization. More precisely, for high self-determined individuals, we suggest that there is an interaction between job demands and job control: high job demands should be positively associated with emotional exhaustion and depersonalization when job control is low. On the other hand, when job control is high, the links between job demands
and the two dimensions of burnout should be insignificant. For low self-determined individuals, we also suggest that there is an interaction between job demands and job control: high job demands should be positively associated with emotional exhaustion and depersonalization when job control is high. However, when job control is low, the relations between job demands and the
two dimensions of burnout should be insignificant.
The second objective of the study is to assess the role of self-determination as a moderator of the demand-control relation in the prediction of learning-oriented outcomes (i.e., personal accomplishment). This variable, which reflects a sense of self-efficacy and productivity at work (Maslach, 1998), was examined in studies that tested the hypothesis of active learning (see Dollard et al., 2000). The second hypothesis postulates that there is a three-way interaction between job demands, job control and self-determination in order to predict personal accomplishment. More specifically, for high self-determined individuals, we postulate that high job demands should be
positively associated with personal accomplishment when job control is high. However, when job control is low, job demands should not be associated with personal accomplishment. For low self- determined individuals, we suggest that a high level of job demands should be positively
associated with personal accomplishment when job control is low. However, when job control is high, job demands should not be associated with personal accomplishment.
Method Participants
Participants were 398 university professors (280 males and 116 females, 2 without gender identification) from a large French-Canadian university (i.e., 1313 professors). The response rate was 30.3. Age ranged from 29 to 72 (mean = 48.89, S.D. = 8.37) and years of experience as university professors ranged from 1 to 43 (mean = 17.61, S.D. = 9.55). Finally, 211 participants held the rank of professor, 120 the rank of associate professor and 35 the rank of assistant professor (32 held other ranks).
Procedure
Data from this study came from a project on university professors’ psychological well- being. A questionnaire was mailed to 1313 professors. In order to guarantee anonymity, participants were asked to mail back the questionnaire. To increase the participation rate, an E- mail message was sent to the professors two weeks after the initial mailing and a telephone follow-up was conducted three weeks after the mailing.
Measures
The questionnaire included the four following measures: 1) job demands, 2) job control, 3) self-determined work motivation, 4) and burnout.
Job demands. Job demands was assessed by a relatively broad range of potential work stressors, namely work overload, role ambiguity, role conflict, and stress related to research activities. To reduce the likelihood of a Type II error (to which tests of interaction are
susceptible), the measure of job demands did not incorporate affective elements. As argued by Wall et al., (1996), the inclusion of an affective element within the independent variable builds in spurious main effects if the dependent variable is also an affective evaluation, such as strain.
Work overload was assessed by ten-items taken from the Occupational Stress Inventory-R (Osipow, 1995). Items focused on qualitative and quantitative demanding aspects of the job (e.g., “I work with time constraints”; a =.82). Items were scored on a five-point scale ranging from 1 (rarely or never) to 5 (constantly or always).
Role ambiguity and role conflict were assessed, respectively, by four-items and eight-items developed by Rizzo, House, & Lirtzman (1970). Role ambiguity concerns the lack of adequate information to do the job adequately (e.g., “I know exactly what is expected of me” (reverse scoring); a =.69). Role conflict concerns the multiplicity of roles and occurred when conflicting demands at the job had to be met (e.g., “I have to do things that should be done differently”; a = .80). Participants used a seven-point scale ranging from 1 (definitively false) to 7 (definitively true). Two original items from the role ambiguity subscale were, a priori, eliminated because they were not relevant to the academic setting.
Work-related research was assessed by five-items taken from Singh, Kim, and Mishra (1998). Each item had a five-point scale ranging from 1 (rarely or never) to 5 (constantly or always). The items assessed the extent to which the participants perceived stress-related research (e.g., “I feel pressured to find funds to publish my studies and research”; a =. 77).
were computed to form a global score of job demands. The Cronbach’s alpha value for this total score was .88.
Job control. Many tests of the Job Demand-Control model have used measures of job control that are confounded with other concepts such as skill variety or job complexity (Ganster,
1989). However, in the present study, we used three items derived from the French version (Brisson, Blanchette, Guimont, Dion, Moisan, & Vezina, 1998) of the Job Content Questionnaire (JCQ; Karasek, 1985). These items concerned opportunities for control and decision and therefore assessed job control per se. This measure is composed of the three following items: "My job allows me to make a lot of decisions on my own” “I have a lot to say about what happens on my job,” and “In my job, I have very little freedom to decide how I work” (reverse scored). Items are
answered on a Likert-type scale ranging from 1 (strongly disagree) to 4 (strongly agree). A mean score was calculated from the three items. The Cronbach’s alpha value for this measure was .80.
Self-determined work motivation. Self-determined work motivation was assessed by the short version of the Blais Work Motivation Inventory (Blais et al., 1993). The short version is made up of 18 items that assess six motivational dimensions (three items per dimension). Two subscales assessed intrinsic motivation: toward accomplishment (e.g., “Because I experience satisfaction when my job provides me with interesting challenges”; a = .76), and towards knowledge (e.g., “Because I experience pleasure when learning new things”; a = .86). Thre subscales assessed types of extrinsic motivation: identified regulation (e.g., “Because this is tW% type of work that I prefer regarding my career aspirations”; a =.57), introjected regulation (e.g., “Because I absolutely want to be good and if I’m not, I’ll be disappointed”; a = .83), and external
regulation (e.g., “For the salary”; a = .77). One subscale assessed amotivation (e.g., “I don’t know, I don’t think that I have what it takes to do this job successfully”; a = .77). Each item represents a possible reason for working. Items were scored on a seven-point scale ranging from 1 (do not agree at all) to 7 (agree completely).
Following the procedure commonly used in the self-determination literature (e.g., Blais, Sabourin, Boucher, & Vallerand, 1990; Grolnick & Ryan, 1987; Guay, Boggiano, & Vallerand, 2001), scores of the six subscales were used to compute a self-determination indicator. For the present study, we used the following formula: [(2* (IM knowledge + IM accomplishment)/2 + 1 * identified regulation) - (1* introjected regulation + 1 *external regulation)/2 + 2* amotivation))]. Positive scores (higher than zero) indicated self-determined motivation whereas negative scores (below zero) revealed non-self-determined motivation.
Burnout. The French version (Dion & Tessier, 1994) of the Maslach Burnout Inventory (MBI; Maslach & Jackson, 1986) was used to measure emotional exhaustion, depersonalization, and personal accomplishment. Emotional exhaustion was measured via seven of the nine items of the original version. Two of the original items that focused on working with people were
eliminated (i.e., "Working with people directly puts too much stress on me" and "working with people all day is really a strain for me”) because few university professors face this kind of stress given the individualistic nature of their tasks. Responses to all items were scored on a seven-point scale ranging from 0 (never) to 6 (every day) (e.g., “I feel emotionally drained from my work”). The Cronbach’s alpha value for this measure was .88.
Five items were used to assess depersonalization. Responses to all items were scored on a seven-point scale ranging from 0 (never) to 6 (every day) (e.g., “I’ve become more callous toward people since I took this job”). The Cronbach’s alpha value for this measure was .78.
Personal accomplishment was measured by 8 items. Responses to all items were scored on a seven-point scale ranging from 0 (never) to 6 (every day), (e.g., “I have accomplished many worthwhile things in this job”). The Cronbach’s alpha value for this measure was .75.
Results Preliminary analyses
Correlations between all variables are presented in Table 1. All predictors (job demands, job control and self-determined work motivation) were significantly correlated with emotional
exhaustion, depersonalization and personal accomplishment. As expected, job demands was positively related to emotional exhaustion (r = .62, p<001), and to depersonalization (r = .38, 2<.001) and negatively to personal accomplishment (r = -.12, p <05). Both job control and self- determined motivation were related positively to personal accomplishment (r = .23, p<001; r = .35, p<001, respectively) and negatively to emotional exhaustion (r = -.41, pc.001; r = -.48, p<.001, respectively), and depersonalization (r = -.26, p<001; r = -.35, pc.001, respectively). Regression analyses
Hierarchical moderated multiple regression analysis (Cohen & Cohen, 1983) was used to test main and interaction effects of job demands, job control and self-determined motivation on the three burnout dimensions (i.e., emotional exhaustion, depersonalization, and personal accomplishment). With this approach, the incremental variance (AR2) accounted for by the interaction term reflects the effect size of the interaction. Three sets of regression analyses were performed to test our two hypotheses. The first regression analysis tested our first hypothesis based on emotional exhaustion. The second regression analysis tested also our first hypothesis based on depersonalization. The third regression analysis tested our second hypothesis based on personal accomplishment. For each analysis, the predictor variables were entered within four
successive steps. In the first step, gender, age, rank and experience were entered to control for potential confounding effects in the academic setting (see Byrne 1991, 1999). In the second step, main effects (job demands, control and motivation) were entered. In the third step, three two-way interaction terms (i.e., demands * control, demands * self-determination, control * self-
determination) were entered. Finally, in the fourth step, the three-way interaction term (i.e., demands * control * self-determination) was entered. To avoid multicollinearity between the predictors and the interaction terms, interactions terms were computed using variables
standardized scores (Aiken & West, 1991; Kline, 1998).
First analysis. Results of this analysis are presented in Table 2. Results from the first step indicated that three of the four control variables were significant and accounted for 12 % of variance in emotional exhaustion. Results of the second step indicated that job demands, job control and self-determined work motivation were all significantly related to emotional exhaustion (AR2 = .37). Results of the third step indicated that only demand-control interaction was
significant, but did not account for a significant percentage of incremental variance. Results of the fourth step indicated that the three-way interaction terms explained 1% of additional variance in emotional exhaustion (p< .05).
In order to interpret this interaction effect, which involved continuous variables, simple slopes were derived for high (+1 SD), and low levels (-1 SD) of the moderator, self-determined work motivation (Aiken & West, 1991). The results, which are shown in Figure 1, indicate that job demands was very positively associated with emotional exhaustion for high self-determined
employees with low job control (ß = .66, p < .001), but there was an attenuated association
between job demands and emotional exhaustion for high self-determined individuals with high job control (ß = .32, p < .001). Job demands was also positively related to emotional exhaustion for
low self-determined employees with both high (ß = .36, p < .001) and low job control (ß = .46, p < .001). These results thus indicated that job control buffers the negative effects of job demands on emotional exhaustion for high self-determined employees.
Second analysis. Results for the depersonalization variable are presented in Table 3. Results from the first step indicated that none of the four control variables was significant. Results of the second step indicated that job demands, job control and self-determined work motivation were all significant and accounted for 22% of variance in depersonalization. Results of the third step indicated that demands-control interaction and demands-self-determination interaction were significant, and accounted for a significant 3% of incremental variance. Results of the fourth step indicated that the three-way interaction term explained a significant portion of the variance (i.e., 1%) in depersonalization (p < .05).
Figure 3 shows that job demands was very positively associated with depersonalization for high self-determined employees with low job control (ß = .56, p < .001). Interestingly, job
demands was not significantly related to depersonalization (ß = .06, ns.), for high self-determined individuals with high job control. In addition, job demands was positively related to
depersonalization for low self-determined employees with low job control (ß = .46, p < .001) but there was an attenuated association for low self-determined employees with high job control (ß = .29, p < .01). The results revealed that job control buffers the negative effects of job demands on depersonalization for high self-determined employees and also, but to a lesser extent, for less self- determined employees.
Third analysis. Results for personal accomplishment are presented in Table 4. Results from the first step indicated that none of the four control variables were significant. Results of the second step indicated that job demand was non-significant, but that job control and
self-determined motivation were significant and accounted for 12% of variance in personal
accomplishment. Results of the third step indicated that none of two-way interaction terms was significant. Results of the fourth step indicated that the three-way interaction term explained a significant portion of the variance (i.e., 1%) in personal accomplishment (p < .05)1.
Figure 4 reveals the significant interactive effect of job demands, job control, and self- determined motivation for personal accomplishment. As expected, job demands was positively associated with the level of personal accomplishment for high self-determined employees with high job control (ß = .20, p < .05). Job demands was not related to personal accomplishment for high self-determined employees with low job control (ß = .02, ns) and low self-determined employees with high job control (ß = -.06, ns). In addition, job demands was positively, but not significantly, associated with personal accomplishment for low self-determined employees with low job control (ß = .09, ns). These results thus revealed that job control buffers the negative effects of job demands on personal accomplishment for high self-determined individuals.
Discussion
The aim of this study was to examine the individual differences inherent to the use of job control in a high job demands context in order to predict burnout. We examined the interactions that exist between job demands, job control and employee personality, i.e. the employee’s degree of self-determined work motivation, in order to predict each dimension of burnout (i.e., emotional exhaustion, depersonalization and personal accomplishment).
The first objective of the study was to assess the role of self-determined work motivation as a variable that moderates the demands-control relation in the prediction of emotional exhaustion and depersonalization. It was hypothesized that self-determined work motivation plays a
emotional exhaustion, job control reduces the unhealthy effects of job demands for high self- determined individuals. However, for low self-determined individuals, the results do not
demonstrate an interactive effect between job control and job demands, but rather a direct effect of job demands on emotional exhaustion and depersonalization. Job demands, however, is positively
associated with emotional exhaustion, independently of the level of job control enjoyed by low self-determined employees. As regards depersonalization, the results indicate that job control considerably moderates the unhealthy effects of job demands for high self-determined employees. Job control also moderates job demands for low self-determined employees, but to a lesser extent. Thus, these results partly support our first hypothesis. Following O’Connor and Vallerand (1994), we hypothesized that low self-determined individuals would find it easier to adjust to a work environment offering few opportunities to make choices and decisions. O’Connor and Vallerand (1994) had demonstrated that non self-determined individuals adjusted better psychologically to a controlling environment that offers few opportunities for choice and autonomy. All in all, the results obtained support the hypothesis of job strain for high self-determined individuals in the prediction of two negative components of burnout, and for low self-determined individuals in the prediction of depersonalization.
The second objective of the study was to assess the role of work self-determination as a moderator of the demands-control relation in the prediction of personal accomplishment at work. It was hypothesized that self-determination moderates the relation between job demands and job control. In fact, for high self-determined individuals, a high level of job demands is positively associated with personal accomplishment at work when they have a high level of job control, but this is not the case when job control is low. For low self-determined individuals, job demands is not related to personal accomplishment even when employees enjoy a high level of job control.
As for low self-determined employees, a low level of job control was positively associated, but not significantly so, with personal accomplishment. However, for high self-determined
individuals, our results support the hypothesis of the development of learning-oriented outcomes of the Job Demand-Control model (Karasek & Theorell, 1990). To the best of our knowledge, this study is the first to have demonstrated a three-way interactive effect between job demands, job control and individual characteristics in the prediction of active learning.
Theoretical Implications
The results observed underline a limitation of the Job Demand-Control model (Karasek, 1979; 1998 Karasek & Theorell, 1990), that is, it does not take individual differences into account. Although our results indicate the importance of job control, they also suggest that individual differences are inherent to the use of job control. More precisely, not all people react in the same way to job control offered by work environment. These differences are expressed through both individual adjustment to job demands and their resulting psychological manifestations. In fact, job control appears to be crucial, especially for high self-determined employees. Under the
pressure of demanding work, the job control of the latter tends to reduce emotional exhaustion and depersonalization, in addition to fostering their feeling of personal accomplishment at work. Conversely, when job demands is high and they do not have a level of job control that meets their needs, they become vulnerable to emotional exhaustion and depersonalization without the
opportunity to develop their feeling of personal accomplishment. These results are consistent with the latest advances in the understanding of burnout. In fact, they simultaneously incorporate both individual and organizational aspects rather than considering them separately. The phenomenon of burnout is therefore examined on the basis of the fit between individual needs and
These points on the role of individual differences offer a more nuanced view of the theoretical foundations of the dynamic version of the Job Demand-Control model (Karasek & Theorell, 1990). Karasek and Theorell (1990) propose two mechanisms through which active learning and psychological tension are related. First, based on the concept of learned helplessness (Seligman, 1975), they stipulate that the accumulation of strain at work (i.e., high-strain job) inhibits persons’ capacity to learn, to actively develop their skills and to feel effective at work. Second, they suggest that the cumulative effect of a high level of job demands and job control (i.e., active job) increases individual skills and a sense of control which, in turn, reduces the perception of stress. Our results take these hypotheses a step further. That is, the consideration of individual differences suggests that this dynamic would be plausible only among high self-
determined individuals who have a level of job control that meets their needs. In fact, in a high job demands situation, they alone would have a sense of developing personal accomplishment.
Thus, in line with the assertions of Karasek and Theorell (1990), this feeling could protect them from emotional exhaustion and depersonalization. Conversely, a lower level of emotional exhaustion or depersonalization could foster the development of personal accomplishment. This mechanism is similar to that advanced by Maslach et al. (2001) to explain burnout. These authors suggest that the occurrence of one of the dimensions of burnout (i.e. emotional exhaustion, depersonalization or loss of the feeling of personal accomplishment) can precipitate or diminish the development of another dimension.
Our results also call into question certain postulates of Self-Determination Theory (Deci & Ryan, 1985, 1991), which stipulates that there exists a fundamental need for autonomy that does not vary from one individual to the next. Our analysis has brought out differences inherent to the individual’s capacity to adjust to job control. Although the opportunity to make choices and
decisions is favorable to psychological well-being, this need is more important for high self- determined individuals. Future research should therefore further examine the notion of the fit between individual needs and environmental characteristics. The degree of fit between
willingness (i.e., need) and possibility (i.e. job control) should thus facilitate our understanding of psychological adjustment to work.
T .imitations and future research
Although the results of the present study are interesting and founded on a relevant theoretical model, certain limitations should be borne in mind when interpreting them. A self- administered questionnaire was used to measure job demands and job control. Thus, despite the validity of the scales, independent and dependent variables from the self-administered
questionnaires may give rise to common variance problems. However, as mentioned earlier, this possibility appears to us to be negligible because the overestimation of common variance
generally occurs on the principal effects rather than on interactions between the variables (Wall et al., 1996).
Because of the cross-sectional nature of the study, we were unable to stipulate cause and effect relationships. Indeed, although self-determination is relatively stable, it includes a temporal dimension that is likely to be influenced as a function of the characteristics of work environment and the employee’s experience. According to Deci and Ryan (1985, 1991), self-determination may be determined both by personal and environmental characteristics. For example, a non self- determined individual with job control could develop more self-determined regulations over time. Conversely, a self-determined employee who lacks job control could become less self-determined. Longitudinal studies would provide useful information in this respect.
Finally, our results cannot be generalized to all employees in the labour market because the sample studied is composed entirely of university professors.
Conclusion
The results of this study underline the important influence of work self-determination on burnout. They indicate that motivation plays a major role in an employee’s adjustment to job demands. It can foster or delay the appearance of burnout. Thus, the extent to which individuals experience emotional exhaustion, depersonalization or a feeling of personal accomplishment at work will depend on the fit between the work organization (i.e., job demands and job control) and their personality (i.e. level of self-determination).
Note
1The percentages of the explained variance for all regression equations are within the typical range (i.e., ΔΚ2 = .01-.03) for moderator effects in non-experimental studies (Champoux & Peeters,
Table 1. Correlations among all variables and descriptive statistics Variables M SD 1 2 3 4 5 6 7 8 9 10 1■ Age 48.90 8.37 __ 2. Gender3 1.71 0.46 21*** 3. Professor’s rank 2.50 0.66 59*** ,25*** 4. Experience 16.66 9.48 ,84*** 27*** ,61***
_
5. Job demands 3.20 0.60 -,14** -,22*** -,06 -,20*** 6. Job control 3.35 0.60 ,13** ,14** ,07 ,15** -,36*** 7. Self-determined 10.09 3.70 ,07 ,16** ,16** ,06 -,38*** 33*** work motivation 8. Emotional 3.27 1.26 -,21*** -,23*** -,11* -,23*** ,62*** -,41*** -־48*** exhaustion 9. Depersonalization 2.30 0.98 -.08 .02 -.04 -.07 .38*** -.26*** -,35*** .48*** _ 1 ().Personal 5.24 0.81 ,05 ,11* ,07 ,05 -,14** ,23*** 35*** -,20*** -.33** accomplishment Note: *¡)<05, ** ¡)<01, *** ¡)< .001 a 1 = female; 2 = male.Table 2. Hierarchical multiple regression analyses for emotional exhaustion Variables entered 1 Steps 2 3 4 1. Gender -.13* -.00 -.01 -.01 Age -.23** -.09 -.09 -.09 Professor’s rank .13* .07 .08 .08 Experience -.13 -.10 -.12 -.11 2. Job demands 43*** 44*** 45*** Job control -.16*** -.14** -.16*** Self-determination ־24*** -.25*** _ 27***
3. Job demands * job control -.10* -.12*
Job demands * self-dertermination .05 .05
Job control * self-determination -.04 .02
4. Job demands * job control * self-determination -.12*
R2 ^2*** 4g*** .50 50***
ΔΚ2 ^2*** 32*** .01 .01*
Note. The displayed coefficients in the four columns are standardised beta weights at each step.
Table 3. Hierarchical multiple regression analyses for depersonalization Variables entered 1 2 Steps 3 4 1. Gender .04 .14** .15** .14** Age -.13 -.02 -.03 -.04 Professor’s rank .04 .00 .02 .02 Experience -.05 -.04 -.04 .03 2. Job demands 29*** .32*** 34*** Job control -.13* -.10 -.13* Self-determination -.22*** -.23*** -.26***
3. Job demands * job control -.16** -.19**
Job demands * self-determination -.03 -.04
Job control *self-determination -.19** -.11
4. Job demands * job control * self-determination -.18*
R2 .02 24*** .27** .28***
AR2 .02 .22*** .03** .01*
Note. The displayed coefficients in the four columns are standardised beta weights at each step.
Table 4. Hierarchical multiple regression analyses for personal accomplishment Variables entered 1 2 Steps 3 4 1. Gender .06 .01 .01 .02 Age .13 .08 .09 .10 Professor’s rank .02 -.01 -.01 -.01 Experience -.10 -.06 -.07 -.08 2. Job demands .09 -.08 -.06 Job control .15** .15** -.18** Self-determination .32*** .32*** 34***
3. Job demands * job control .01 .01
Job demands * self-determination .05 .05
Job control * self-determination .07 -.01
4. Job demands * job control * self-determination .17*
R2 .01 24*** .14 25***
AR2 .01 22*** .00 .01*
Note. The displayed coefficients in the four columns are standardised beta weights at each step. *p<.05, **p<.01, ***p^.001
Figure Caption
Figure 1. The Job Demand-Control model (adapted from Karasek, 1979).
Figure 2. Three-way interaction between job demands, job control and self-determined work motivation when predicting emotional exhaustion: Fitted regression equation ± SD from the mean. Figure 3. Three-way interaction between job demands, job control and self-determined work motivation when predicting depersonalization: Fitted regression equation ± SD from the mean. Figure 4. Three-way interaction between job demands, job control and self-determined work motivation when predicting personal accomplishment: Fitted regression equation ± SD from the mean.
Job demands
Low High Learning-oriented
High Job control Low Job strain Lotv-strain /י Aç.tivé High,strain ____________
High self-determined (high job control) High self-determined (low job control) Low self-determined (high job control) Low self-determined (low job control) ך 1 -0,8 0,6 - 0Λ - 0,2- 0 -ס,: -4 ס, 6 ס, 8 ס, -! 1 Job demands -1 Em oti ona l ex ha us ti on
♦-High self-determined (high job control) —B—High self-determined (low job control) —A—Low self-determined (high job control) X Low self-determined (low job control) ר 1 0,8 -0,6 0,4 -0,2 -0 -0,2 ־ -0,4 -0,6 -8 ס, -1 Job demands ! Deper sonali zati on
♦ High self-determined thigh job control) —■—High self-determined (lowjob control) ■ ■A Low -self-determined (high job control) ..-X■ ■Low self-determined (lowjob control) 0,8 0,6 0,4 0,2 0 --0,2 -1 Job demands 1 Pe rs on al ac co m pl is hm en t
Conclusion générale
Ce mémoire a étudié les différences individuelles inhérentes à Γutilisation de l’autonomie décisionnelle dans un travail où la demande psychologique est élevée, afín de prédire l’épuisement professionnel. À cette fin, il a vérifié les interactions qui existent entre la demande psychologique, l’autonomie décisionnelle et la personnalité de l’employé, soit son degré d’autodétermination, afin de prédire les trois principales dimensions de l’épuisement professionnel (i.e., l’épuisement
émotionnel, la dépersonnalisation et le sentiment d’accomplissement).
Le premier objectif de l’étude consistait à évaluer le rôle de 1 ’autodétermination comme variable modératrice de la relation demande-autonomie dans la prédiction de l’épuisement
émotionnel et la dépersonnalisation. Tel qu’il a été postulé, T autodétermination au travail modère la relation entre la demande psychologique et l’autonomie décisionnelle. Dans la prédiction de l’épuisement émotionnel, l’autonomie décisionnelle atténue les effets néfastes de la demande psychologique pour les gens fortement autodéterminés. Cependant, pour les gens peu
autodéterminés, les résultats ne démontrent pas d’interaction significative entre l’autonomie décisionnelle et la demande psychologique, mais plutôt un effet direct de la demande
psychologique sur l’épuisement émotionnel et la dépersonnalisation. Ceci étant dit, la demande psychologique est associée positivement à l'épuisement émotionnel, indépendamment du niveau d'autonomie décisionnelle dont bénéficie l’employé peu autodéterminé. En regard à la
dépersonnalisation, les résultats indiquent que l’autonomie décisionnelle modère
considérablement les effets néfastes de la demande psychologique pour les gens fortement
autodéterminés. L’autonomie décisionnelle modère également la demande psychologique mais à un moindre degré pour les gens faiblement autodéterminés. Ces résultats appuient donc,
travail pour les gens ayant une motivation fortement autodéterminée au travail dans la prédiction des deux composantes négatives de Γépuisement professionnel et pour les gens peu
autodéterminés dans la prédiction de la dépersonnalisation.
Le second objectif de l’étude visait à évaluer le rôle de 1 ’autodétermination comme variable modératrice de la relation demande-autonomie dans la prédiction de Γ accomplissement personnel au travail. Tel qu’il a été postulé, 1 ’autodétermination au travail modère la relation entre la demande psychologique et l’autonomie décisionnelle. En effet, pour les gens plus
autodéterminés, un niveau élevé de demande psychologique est associé positivement à
!'accomplissement au travail lorsque l'autonomie décisionnelle est élevée, mais ne l’est pas lorsque l'autonomie décisionnelle est faible. Pour les gens peu autodéterminés, la demande psychologique n’est pas reliée au niveau d'accomplissement au travail même si les gens bénéficient d’un niveau élevé d’autonomie décisionnelle. En ce qui a trait aux employés peu autodéterminés, un faible niveau d’autonomie décisionnelle est positivement associé, mais de façon non significative, à l’accomplissement au travail. Cette étude est donc la première à démontrer un triple effet d’interaction entre la demande psychologique, l’autonomie décisionnelle et les caractéristiques individuelles et ce, dans la prédiction des apprentissages actifs.
En dépit de certaines limites, cette étude contribue à l’avancement des connaissances sur les liens entre !’organisation du travail et l’épuisement professionnel. En corroborant le rôle de l’autonomie décisionnelle, elle souligne qu’il existe des différences individuelles inhérentes à !’utilisation de celle-ci. Ces différences s’exprimeraient autant dans l’adaptation individuelle vis- à-vis de la demande psychologique que dans les manifestations psychologiques qui en découlent. De fait, l’autonomie décisionnelle apparaît cruciale surtout pour les employés ayant une
décisionnelle de ces derniers tend à réduire l’épuisement émotionnel et la dépersonnalisation, en plus de favoriser leur accomplissement au travail. À l’inverse, lorsque la demande est forte et qu’ils ne peuvent pas jouir d’un niveau d’autonomie décisionnelle répondant à leurs besoins, les employés deviennent très vulnérables à l’épuisement émotionnel et à la dépersonnalisation, sans avoir l’occasion de développer leur sentiment d’accomplissement personnel. Ces résultats
rejoignent donc les dernières percées en matière de compréhension de l’épuisement professionnel En effet, ils intègrent les aspects individuels et organisationnels simultanément, plutôt que de les considérer séparément.
En somme, l’étude souligne l’importance de 1 ’autodétermination comme facteur d’influence sur l’épuisement professionnel. Les résultats indiquent que l’orientation
motivationnelle de l’employé est capitale dans son adaptation vis-à-vis de !’organisation de son travail. Ainsi, elle est susceptible de favoriser ou retarder l’apparition de l’épuisement
professionnel. L’individu vivra donc de l’épuisement émotionnel, de la dépersonnalisation ou développera son accomplissement personnel au travail, s’il y a adéquation ou inadéquation entre !’organisation du travail (i.e., la demande psychologique et l’autonomie décisionnelle) et sa personnalité (i.e., l’autodétermination au travail).
Références
Aitken, L. S., & West, S. G. (1991). Multiple Regression: Testing and interpreting interaction. Newbury Park, CA: Sage.
Alfredsson, L., Spetz, C. L., & Theorell, T. (1985). Type of occupation and near-future hospitalization for myocardial infarction and some other diagnoses. International Journal of Epidemiology. 14, 378-388.
Beerh, T. A., Glaser, K. M., Canali, K., & Wallwey, D. A. (2001). Back to basics: Re- examination of Demand-Control Theory of occupational stress. Work and Stress. 15(21. 115-130.
Blais, M. R., Lachance, L., Brière, N. M., Riddle, A. S., & Vallerand, R. J. (1993). L'inventaire des motivations au travail de Blais. Revue Québécoise de Psychologie. 14. 185-215.
Blais, M. R., Hess, U., Bourbonnais, J., Saintonge, 1, & Riddle, A. (1995). Mens sana ad corpus sanum : Un modèle de motivation-stress-santé appliqué au couple et au travail. Santé mentale au Québec. 20(2). 139-162.
Blais, M. R., Sabourin, S., Boucher, C., & Vallerand, R. J. (1990). Toward a motivational model of couple happiness. Journal of Personality and Social Psychology, 59. 1021-1031.
Blix, A. G., Cruise, R. J., Mitchell, B. M., & Blix, G. G. (1994). Occupational stress among university teachers. Educational Research. 36(2). 157-169.
Brisson, C., Blanchette, C., Guimont, C., Dion, G., Moisan, J., & Vézina, M. (1998). Reliability and validity of the French version of 18-item Karasek Job content questionnaire. Work and Stress, 12(4), 322-336.
Byrne, B. M. (1991). Burnout: Investigating the impact of background variables for elementary, intermediate, secondary, and university educators. Teaching and Teacher Education : An International Journal of Research. 7, 197-209.
